Museum objects, including paintings, incorporate in their compositions a wide range of natural organic materials. First there are the materials such as wood, animal and vegetable fibres, and skin products such as leather and parchment, which can play a major structural role. Then there are the minor (in terms of bulk) components such as drying oils, waxes, plant gums, proteins such as albumin and gelatin, and the natural plant resins, which are present as adhesives, varnishes, or binding media for paint. Distinguishing between materials of the first group can usually be done by simple inspection, while determination of generic or species origins, of woods say, normally involves following well-defined routines in microscopy rather than any kind of chemical analysis.
With the minor components, however, it is not normally possible to tell from simple inspection what class of material is involved, and indeed the material may not even, as with paint media, be visible for inspection in any real sense. Relatively simple tests of solubility or chemical properties will quite often serve to answer this first question, though not if mixtures are involved or properties are disguised by the presence of large amounts of inorganic materials. In the latter cases, and when more specific identification is wanted it is necessary to resort to fairly sophisticated chemical and instrumental techniques and ones, moreover, which are capable of giving results with very small samples.
There are essentially no methods of organic analysis applicable to these materials which can be used on the object itself; it is always necessary to remove a sample for study. Furthermore most of the methods will also involve the destruction of this removed sample by dissolving it to separate organic and inorganic components, or to get it into suitable form for the method to be used. Obviously this is a drawback when structured samples, such as those from paintings, are involved. Some tests on cross-sections of such samples can be effected by biochemical staining in favourable cases, but this is more the concern of the microscopist. To get really meaningful results from the following instrumental methods therefore, samples which are as homogeneous as possible are desirable.
The two main areas of analysis for which the Department is equipped are those involving spectrometric methods and the various forms of chromatography. Ultra-violet and visible spectrometry is only rather rarely useful. Few organic materials encountered in old objects retain distinctive ultra-violet spectra that they may originally have had. This is for a very good reason, namely that the possession of such absorption renders the compounds naturally susceptible to photochemical reactions and hence change. With coloured compounds the possession of a visible absorption spectrum is an inevitable attribute and measurement of this is a valuable aid to identification with organic dye- stuffs that can be separated from the substrate and brought into solution in sufficient quantity. This is only rarely possible with samples from paintings which is why the direct measurement of such spectra on thin sections, described elsewhere in this bulletin, has been investigated. Infra-red spectrometry is a basic technique for identification of organic materials but again one which is only rarely useful for samples from paintings. It comes into its own with samples which are not too complex mixtures of disparate chemical types in the first place, and which are not too liable to chemical change with time, and waxes are a good example of such material. Even with these, however, more convincing analyses (especially with mixtures) on smaller samples are obtainable using gas-chromatography.
Gas-chromatography is the most valuable and generally applicable of the chromatographic techniques, and indeed of all the methods available to us. The Scientific Department obtained its first gas-chromatograph (a Pye 'Panchromatograph') in 1962 before anything was known of its usefulness in this field and consequently with some misgivings as to whether it would prove a good investment. Already by the following year its value for paint medium analysis was established and it has been used almost routinely for this since then. A much superior Pye 104 instrument has been in use since 1970 (Fig.8). As mentioned above gas- chromatography is also used for the identification of waxes and indeed for most of the groups of natural materials which have been mentioned. Thus proteins are identified by GLC of derivatives of their constituent amino acids, and the natural resins can also be characterized after suitable derivitisation of their component di- and triterpenes. Resins containing triterpenes (such as dammar and mastic) are a particular challenge because of the high molecular weight (and hence boiling point) of their components. A system of capillary column chromatography of the trimethylsilyl derivatives has been developed for these.
Gas-chromatography identifies natural materials by separating them into their components. The pattern of peaks which results may sometimes be used simply as a characteristic of the material but more often, and more informatively, the chemical identities of the compounds responsible for each peak are ascertained by comparison with known samples and their amounts estimated from the peak sizes. Thus information may be acquired, not only regarding the nature of the raw materials, but also the ways in which they change with time. Some materials change very much with time and produce many new unidentifiable peaks on the chromatograms. It would be most desirable to be able to have fuller information on these and this could, of course, be obtained with the now well developed methods of combined gas-chromatography/mass-spectrometry linked with a computer data-processing unit. The relatively high cost of such an installation has so far prevented our acquiring it.